[Bug]: How to use MVTec datasets for multi-classes training

[zzh805780186]

Describe the bug

when I use MVTec datasets to training model with config.yaml (eg: anomalib/models/padim/config.yaml), I see the 'category' just write one class('bottle'). Does it mean training only with one class?

How do I modify the config file if I want to train the model with multiple categories? (eg:bottle,cable,capsule,carpet...). I have tried to modify it in several ways, as shown in the picture, but nothing works.

Dataset

MVTec

Model

FastFlow

Steps to reproduce the behavior

python tools/train.py --model fastflow --config anomalib/models/fastflow/config.yaml

OS information

OS information:

• OS: [e.g. Ubuntu 20.04] Ubuntu 18
• Python version: [e.g. 3.8.10] Python 3.8.16
• Anomalib version: [e.g. 0.3.6] Anomalib 0.4.0
• PyTorch version: [e.g. 1.9.0]
• CUDA/cuDNN version: [e.g. 11.1]
• GPU models and configuration: [e.g. 2x GeForce RTX 3090]
• Any other relevant information: [e.g. I'm using a custom dataset]

Expected behavior

How to use MVTec datasets for multi-classes training

Screenshots

No response

Pip/GitHub

pip

What version/branch did you use?

No response

Configuration YAML

dataset:
  name: mvtec
  format: mvtec
  path: ./datasets/MVTec
  task: segmentation
  category: [bottle,cable,capsule,carpet,grid,hazelnut,leather,metal_nut]
  #category: bottle
  train_batch_size: 32
  eval_batch_size: 32
  num_workers: 8
  image_size: 256 # dimensions to which images are resized (mandatory) options: [256, 256, 448, 384]
  center_crop: null # dimensions to which images are center-cropped after resizing (optional)
  normalization: imagenet # data distribution to which the images will be normalized: [none, imagenet]
  transform_config:
    train: null
    eval: null
  test_split_mode: from_dir # options: [from_dir, synthetic]
  test_split_ratio: 0.2 # fraction of train images held out testing (usage depends on test_split_mode)
  val_split_mode: same_as_test # options: [same_as_test, from_test, synthetic]
  val_split_ratio: 0.5 # fraction of train/test images held out for validation (usage depends on val_split_mode)
  tiling:
    apply: false
    tile_size: null
    stride: null
    remove_border_count: 0
    use_random_tiling: False
    random_tile_count: 16

model:
  name: fastflow
  backbone: resnet18 # options: [resnet18, wide_resnet50_2, cait_m48_448, deit_base_distilled_patch16_384]
  pre_trained: true
  flow_steps: 8 # options: [8, 8, 20, 20] - for each supported backbone
  hidden_ratio: 1.0 # options: [1.0, 1.0, 0.16, 0.16] - for each supported backbone
  conv3x3_only: True # options: [True, False, False, False] - for each supported backbone
  lr: 0.001
  weight_decay: 0.00001
  early_stopping:
    patience: 3
    metric: pixel_AUROC
    mode: max
  normalization_method: min_max # options: [null, min_max, cdf]

metrics:
  image:
    - F1Score
    - AUROC
  pixel:
    - F1Score
    - AUROC
  threshold:
    method: adaptive #options: [adaptive, manual]
    manual_image: null
    manual_pixel: null

visualization:
  show_images: False # show images on the screen
  save_images: True # save images to the file system
  log_images: True # log images to the available loggers (if any)
  image_save_path: null # path to which images will be saved
  mode: full # options: ["full", "simple"]

project:
  seed: 42
  path: ./results

logging:
  logger: [] # options: [comet, tensorboard, wandb, csv] or combinations.
  log_graph: false # Logs the model graph to respective logger.

optimization:
  export_mode: null #options: onnx, openvino

# PL Trainer Args. Don't add extra parameter here.
trainer:
  enable_checkpointing: true
  default_root_dir: null
  gradient_clip_val: 0
  gradient_clip_algorithm: norm
  num_nodes: 1
  devices: 1
  enable_progress_bar: true
  overfit_batches: 0.0
  track_grad_norm: -1
  check_val_every_n_epoch: 1 # Don't validate before extracting features.
  fast_dev_run: false
  accumulate_grad_batches: 1
  max_epochs: 500
  min_epochs: null
  max_steps: -1
  min_steps: null
  max_time: null
  limit_train_batches: 1.0
  limit_val_batches: 1.0
  limit_test_batches: 1.0
  limit_predict_batches: 1.0
  val_check_interval: 1.0 # Don't validate before extracting features.
  log_every_n_steps: 50
  accelerator: auto # <"cpu", "gpu", "tpu", "ipu", "hpu", "auto">
  strategy: null
  sync_batchnorm: false
  precision: 32
  enable_model_summary: true
  num_sanity_val_steps: 0
  profiler: null
  benchmark: false
  deterministic: false
  reload_dataloaders_every_n_epochs: 0
  auto_lr_find: false
  replace_sampler_ddp: true
  detect_anomaly: false
  auto_scale_batch_size: false
  plugins: null
  move_metrics_to_cpu: false
  multiple_trainloader_mode: max_size_cycle

Logs

(anomalib_env) zzh@lc:~/Anomaly_Detection/anomalib-main$ python tools/train.py --model fastflow --config anomalib/models/fastflow/config.yaml
Traceback (most recent call last):
  File "tools/train.py", line 75, in <module>
    train()
  File "tools/train.py", line 50, in train
    config = get_configurable_parameters(model_name=args.model, config_path=args.config)
  File "/home/zhaozihao/anaconda3/envs/anomalib_env/lib/python3.8/site-packages/anomalib/config/config.py", line 265, in get_configurable_parameters
    project_path = project_path / config.dataset.category
TypeError: unsupported operand type(s) for /: 'PosixPath' and 'ListConfig'

Code of Conduct

• I agree to follow this project's Code of Conduct

[alexriedel1]

You create a new folder (for example "all defects"), put all the anomaly images inside and now you have a new category with all classes :)

[KDD2018]

You create a new folder (for example "all defects"), put all the anomaly images inside and now you have a new category with all classes :)

Hi @alexriedel1, I have a question. If I put 'all defects' in a folder, how do I get specific one of all defects in inference phase?